Warping and bowing of marble panels

Microsoft Word - Micro project.docxMicro project Science and engineering of natural stone and glass
autumn semester 2015
completed by:
Stephanie Keine
Zurich, 18 December 15
2   Problem statement ................................................................................................. 2  
3   Studies ..................................................................................................................... 3   3.1   First considerations ............................................................................................ 3   3.2   The TEAM project .............................................................................................. 3   3.3   Thermal cycle tests ............................................................................................ 3   3.4   Terrestrial laser .................................................................................................. 4  
4   Solution approaches .............................................................................................. 5   4.1   Climatic conditions ............................................................................................. 5   4.2   Local installation ................................................................................................ 5   4.3   Stone properties ................................................................................................ 5  
5   Conclusion .............................................................................................................. 6  
1 Introduction
Thin marble panels have become very popular as building claddings recently. Although
looking very good in many buildings, problems have occurred in some façades. The
panels tend to bow, convex or concave, and decrease highly in strength.
At the Finlandia hall (picture 1), a large concert and congress hall in Finland fully
cladded with Italian Carrara marble panels, the marble had deteriorated twice in a
period of just 25 years. The marble panels started to bow concave (picture 2) on all
façades just a few years after completion [Royer-Carfagni 1999]. In 1991, 20 years
after construction, the first panel fell off and a vast loss of strength was recorded. The
cladding was replaced in 1997 with panels of the same marble but in a smaller size and
on a flexible anchoring system. However, just six month after the replacement the
panels bowed again, this time mainly convex, and showed a decrease in flexural
strength. The renovation costs have been summed up to 4.7 Million $ [Grelk et al.
2006].
Picture 1: Finlandia hall, Finland1 Picture 2: bowing marble 2
As similar incidents have been recorded all over the world much research has been
done on this topic and a large project, the TEAM project, was set up to find the sources
of the problem as well as a way of measurement to predict the deterioration probability
of certain types of marble. This work shall outline the research and give a short
introduction to potential solutions.
2
2 Problem statement
Bowing marble has already been registered at grave stones in former times [Grimm
1999], but this was not considered to be a general problem. However, with the
increasing popularity of marble as a building cladding, incidents like the one described
above have been reported from all over the world. Most events took place in Europe or
North America, where marble claddings are used very commonly, but also buildings in
Libya, Lebanon and Cuba had problems with bowing marble panels [Grelk et al. 2007].
The majority of the bowing marble was Italian Carrara marble, but as it is the most
widely used marble, there are also many buildings where the Carrara marble did not
show any changes. In Addition, other marbles from America, Norway or Portugal have
shown a similar bowing behaviour [Grelk et al. 2006]. Still of two buildings in one town
cladded with the same marble, one may bow whereas the other one does not have any
problems.
All panel sizes from small to very large as well as panels of various thicknesses have
shown bowing. It has even been reported from thick marble plates of 8 to 9 cm. Also
different anchoring systems have not helped to prevent the marble from bowing, the
anchoring points mostly being the point of failure.
The most important problem is though that once the panels bowed and the correlated
deterioration has started the phenomena can not be reversed, no mending method is
known by now. So the only solution is a full replacement of the marble cladding
implicating all uncertainties of a new build façade mentioned above as well as high
costs. Therefore there is a high interest in stating the reasons of marble bowing.
3
3.1 First considerations
Research on marble has started very early in 1919 by D. W. Kessler [Kessler 1991]
who examined the expansion and contradiction of marble masonry stones as well as
the bowing of marble grave stones. He already referred the bowing phenomena to
repeated heating. Later studies, e.g. [Bortz et al. 1988] who analysed problems of
different veneer stones, also included a change in moisture as a reason for
deterioration. A first larger study especially on marble panels modelled the thermal
expansion and examined an influence of microcracks [Siegesmund et al. 1999].
3.2 The TEAM project
A big study was finished by the TEAM project (Testing and Assessment of marble and
limestone) in 2007 collecting all former research on marble bowing and analysing the
marble panel claddings of more than 200 buildings [Schouenborg et al. 2007]. Diverse
existing hypotheses were collected and determined in laboratory studies, leading to
variations in temperature and moisture as well as the microstructure being considered
as the most important factors. Furthermore they investigated the “bow-meter”, a tool to
measure the bowing amplitudes with high precision and repeatable. So it can also be
used to perform long time studies e.g. to assess the bowing rate evolution which can
give information about the durability of an already bowed façade. The bowing rate
shows a good correlation with the long-term residual strain, but not with the loss of
strength; a loss of strength could occur
Furthermore the TEAM project developed the adjacent grain analysis (AGA), a
combination of grain size and grain shape analysis, which gives a first idea about the
suitability of a marble.
3.3 Thermal cycle tests
Further laboratory studies to the influences of temperature and moisture were
performed comparing three different marble types taken from three different buildings
in [Siegesmund et al. 2008]. All three considered marbles showed bowing as well as
deterioration and cracks. Two of the marbles, a Peccia marble and a Rosa Estremoz
marble, bowed mainly convex, the third one, a Carrara marble, however, bowed mainly
concave. Differences in bowing intensity were registered for different heights and
façade orientations. In the laboratory tests mercury porosimetry was used to
investigate changes in porosity and pore size distribution which could be related to the
bowing intensity. In Addition, panels were exposed to heating cycles with temperature
4
changes in wet and dry conditions. It was shown that a dry condition and a high
temperature difference only leads to small residual strains whereas a wet condition and
a high temperature difference leads to a much higher residual strain for all marble
types considered. However, the Carrara marble experienced much higher residual
strains than the others. The Peccia marble showed very different bowing depending on
the cutting direction.
A further study [Luque et al. 2010] compared two calcitic and one dolomitic Macaea
marbles from the same geographic area in Spain. The marble samples were again
exposed to first dry and later wet thermal cycles and measurements were done using
optical microscopy, ultrasound test and hot-stage ESEM imaging. The optical
microscopy showed an equidimensional and homogeneous crystal system for the
White and Yellow Macael, but a non-equidimensional system for the Tranco Macael. In
the thermal cycles it was confirmed that a wet condition and a high temperature
difference leads to a higher residual strain which, as the panels are fixed, leads to
bowing. Again, large differences were received for the different types of marble. The
dolomitic Yellow Macael showed a very low increase in the residual strain, the Tranco
Macael a medium increase and the White Macael a very high one. Furthermore,
microcracks are observed for all three types during heating, but while they closed again
during cooling at the Tranco Macael, they stayed open for the White and Yellow
Macael .
3.4 Terrestrial laser
A proceeding technique to the bow-meter was found in the terrestrial laser [Al-Neshawy
et al. 2010]. It can be used to measure the deterioration and the damage dimension on
site as the laser scanning allows a non-invasive view into the marble. Furthermore the
bowing behaviour can be described by fitting a quadratic polynomial curve to the
scanned points. As the terrestrial laser in comparison with the “bow-meter” shows quite
similar results for convex bowed panels, but different ones for concave bowed panels it
should only by used as an additional method.
5
4 Solution approaches
4.1 Climatic conditions
The climate seems to have the highest influence on bowing. This is, however, not
determined by a specific region, but by the existence of moisture and temperature
changes. As seen in the field observations as well as in the laboratory studies the
higher the temperature changes are, the more likely the panels are to bow. This is also
important for different façade orientations as e.g. the East and West façade may not
experience the highest temperatures, but may have a higher difference in temperature
over a day or year. Also building density, shadowing and height can have a high
influence on the actual temperatures, so this could be a reason why buildings in the
same region with the same marble sometimes bow and sometimes do not.
Although the climatic conditions are too versatile to face the problem the TEAM project
found a solution to deal with it: by applying an impregnation to the panels moisture can
be retarded intruding the marble [Schouenborg et al. 2007]. Microcrystalline wax has
been considered to be a good impregnation for marble panels. The impregnation can
also be applied to already bowed panel, although the deterioration level should be
checked.
4.2 Local installation
The local installation is found not to be a main topic in the prevention of marble bowing.
Although some anchoring systems work better than others, as an undercut anchoring
system requires less strength, the anchoring system will not stop the bowing
phenomena. Thicker panels tend to bow slower than thinner panels, however, they still
are too problematic as there is no safe thickness [Schouenborg et al. 2007].
4.3 Stone properties
The microstructure is the second important part when investigating the bowing
behaviour. As shown in [Siegesmund et al. 2008] the degree of bowing depends on the
cutting direction, so the bowing depends on the fabric. In Addition, the Macael marbles
examined in [Luque et al. 2010] differ in the microstructure. A non-equidimensional
microstructure can rearrange cracks formed by heat, so a panel with this microstructure
enlarged by thermal changes can come back to an initial state more easily. So a
irregular microstructure of the marble seems to preserve the marble from bowing rather
than a clear microstructure. Furthermore dolomitic marbles seem to have less
problems with bowing.
5 Conclusion
Summing up, the investigations have shown that some types of marble are more likely
to bow than others. So marbles with a non-equidimensional structure should be
preferred. However, as all marble blocks even from the same type and area can
behave very differently, tests on flexural strength and thermal extension should be
executed before a specific marble is used. To prevent the panels from moisture
intrusion an impregnation, preferably a microcrystalline wax, should be applied. When
this is done the anchoring system and panel size is free to choose.
To monitor the behaviour of the marble panels frequent measurement with the bow-
meter and the terrestrial laser should be made. The terrestrial laser might even
recognize first deteriorations or fading of the impregnation before bowing.
7
References
[Al-Neshawy et al. 2010] F. Al-Neshawy, J. Piironen, S. Peltola, A. Erving, N. Heiska, M. Nuikka, J. Puttonen, “Measuring the bowing of marble panels in building façades using terrestrial laser scanning technology” in Journal of Information Technology in Construction (ITcon), Vol. 15, pp. 64-74, 2010.
[Bortz et al. 1988] S.A. Bortz, B. Erlin, C.B. Monk Jr., “Some field problems with thin
veneer building stones” in New stone technology, design and
construction for exterior wall system. ASTM Special Technical
Publication, Vol. 996, pp.11-31, 1988.
[Grelk et al. 2006] B. Grelk, B. Schouenborg, K. Malaga, “Deterioration of Thin Marble Cladding – a major international study” in “Discovering Stone”, Issue 9, pp. 22-29, 2006.
[Grelk et al. 2007] B. Grelk, C. Christiansen, B. Schouenborg, K. Malaga, “Durabilty of Marble Cladding – A Comprehensive Literature Review” in Journal of ASTM International, Vol. 4, No. 4, 2007.
[Grimm 1999] W. Grimm, “Beobachtungen und Überlegungen zur Verformung von Marmorprojekten durch Gefügeauflockerung” in Zeitschrift der deutschen geologischen Gesellschaft” Vol. 150, Nr. 2, pp. 195-236, 1999.
[Kessler 1991] D.W. Kessler, “Physical and chemical tests on the commercial marbles of the united states” in Technologic Papers of the Bureau of Standards, No. 123, 1919.
[Luque et al. 2010] A. Luque, E. Ruiz-Agudo, G. Cultrone, E. Sebastián, S. Siegesmund, “Direct observation of microcrack development in marble caused by thermal weathering”, in Environmental Earth Sciences, Vol. 62, pp. 1375-1386, 2011.
[Royer-Carfagni 1999] G. Royer-Carfagni, “Some considerations on the warping of marble façades: the example of Alvar Aalto’s Finland Hall in Helsinki”, in Construction and Building Materials”, Vol. 13, pp. 449-457, 1999.
[Schouenborg et al. 2007] B. Schouenborg, B. Grelk, K. Malaga, “Testing and Assesment of Marble and Limestone (TEAM – Important Results from a Large European Research Project on Cladding Panels” in Jornal of ASTM international, Vol. 4, No. 5, 2007.